Michael Lewyn's article Sprawl in Europe and America attempts to demonstrate that suburbanization (pejoratively called "sprawl") is not, as Robert Bruegmann suggests, a predictable result of increasing wealth. He further indicates that suburbanization occurs only to a "limited extent" in Europe. Bruegmann's authoritative volume, Sprawl: A Compact History shows that suburbanization has occurred throughout history and continues to occur virtually wherever people become more affluent.
Lewyn refers to Bruegmann's view as the "inevitability theory of sprawl" and proceeds in his attempt to refute it by not talking about suburbanization. Lewyn largely limits his analysis to post 1990 Europe, and makes so much of recent central city growth in Europe that readers could be misled into thinking that people are abandoning the suburbs in a rush to "return" to the central city. Lewyn's contribution confuses the issue.
Lewyn relies on a review of public transport ridership and central city (excluding suburban) population to prove his thesis. Neither public transport nor central city population is an indicator of suburbanization. Suburbanization is measured by the physical expansion of urban areas outside previous urban footprints. Using transit data could lead an analyst, for example, to conclude the Boston has a smaller urban footprint than Los Angeles, which is evidenced by its higher relative transit use. In fact, United States Bureau of the Census data indicates that the Boston urban area had less than one-third the population of the Los Angeles urban area in 2000 and was, as a result, considerably more suburban in relative terms than Los Angeles.
Even so, Lewyn claims that public transport ridership is increasing in European urban areas. This may or may not be true. His data source does not provide such information. Euro stat (European Union) data does not differentiate between urban and intercity travel. Thus, there is no readily available national or European Union data because both bus and suburban (commuter) rail cannot be separated from the overall total urban and intercity figures.
A genuine comparison of automobile-oriented suburbanization in the United States and Europe must begin before 1990. By 1990, substantial automobile-oriented suburbanization had already occurred in Europe. Virtually all of the large central cities that did not expand their boundaries and which were fully developed 25 years earlier had fallen from their population peaks (as they had in the United States). For example, the core city of Copenhagen, often cited by American planners for its virtues, lost 39% of its population from 1950 to 1990, a figure not much less than the approximately 45% losses in Cleveland and Detroit. The ville de Paris lost approximately one-quarter of its population.
Suburbanization can only be analyzed by looking at suburbanization. It is not sufficient to rely principally on a review of a few areas going back to 1990, as Lewyn does. In fact, suburbanization can by no means be considered "limited" in Europe. From the mid-1960s to the beginning of the new millennium, the metropolitan areas of Europe with populations over 1 million gained 24 million residents. The suburbs captured 27 million of them-yes, just like some American central cities, European central cities sustained losses-to the tune of more than 3 million residents. The limited data on urban area density indicates a decline of 50% in central city population density from the 1960s to 2000.
A complete analysis would also take into consideration the very different urban forms that existed in Europe and the United States even on the eve of automobility. At the beginning of the 20th century, Europe's largest urban areas were far more compact than in the United States, principally because they contained large segments that had been developed before public transport, much less the car. Few pre-19th century urban cores were ever built in the U.S., and they were small. The far more considerable pre-19th century cores continue to exist in Europe to this day. Building of such cores stopped many decades ago.
Affluence must also be considered. Modern middle-income household affluence came earlier in the United States and was further postponed in Europe by a decades-long recovery from World War II. The result is that automobile-oriented suburbanization came earlier in the United States, because Americans were affluent enough to buy automobiles long before Europeans. American 1930 household rates of automobile ownership were not reached in the most affluent Western European nations until the 1970s. The United States has been building automobile-oriented suburbs for decades longer than the Europeans, who nonetheless have built them with a vengeance.
Population growth has also been much faster in the United States. This is important, because nearly all growth in the United States and Europe has been in suburban areas. The United States has had many more years and many more people to settle in suburban areas, so it is no wonder that U.S. suburbs are more expansive than their European counterparts.
Lewyn rightly notes that some European central cities have gained population since 1990, something that has also occurred in some U.S. central cities. But the significance of this should not be exaggerated. Inner London, which has experienced Europe's greatest recent central city growth, has made up less than one-quarter of its population loss since 1911. The ville de Paris has made up less than 10% of its population loss since 1954.
Lewyn further claims that some European central cities have grown faster than their metropolitan areas. While there are a few cases of this, Lewyn's analysis relies upon inconsistent and inappropriate data. In some cases, Lewyn uses urban agglomeration (urban area) data and in other cases, metropolitan area data. There is a difference. Urban agglomerations are urbanized areas or urban footprints and contain no rural territory. Metropolitan areas are labor markets and always include rural areas. The two terms may not be used interchangeably, and considerably different results can be obtained from analyzing one or the other. In the most extreme case (London), there is at least a 4 million population difference between the urban area and the metropolitan area.
A particularly egregious example is Lewyn's analysis of Vienna. Not finding readily available metropolitan area data, Lewyn uses the entire nation of Austria as a surrogate for the Vienna metropolitan area. In fact, Austria is far larger than any reasonable definition of a metropolitan area and cannot be appropriately used as a metropolitan surrogate. One might as legitimately use the state of Indiana as the Indianapolis metropolitan area (Indiana covers only marginally more land area than Austria).
Further, the European suburbs are far less dense than the central cities from which people exited. European suburbs average a population of 6,600 per square mile, while the central cities average 18,800. There is no doubt that European suburbs are more dense than U.S. suburbs (2,700 per square mile), but they are still suburbs and they are automobile-oriented.
Moreover, it would be wrong to assume that core city population growth represents a "return to the city" or that the suburbanization trends that have been under way for decades have been reversed, much less stopped. This is illustrated by Munich and Vienna, two of Lewyn's cases.
Munich commuting patterns illustrate that suburbanization continues apace. Between 1998 and 2006, the city of Munich accounted for 12%of the new commuting to jobs in the Munich metropolitan region. The suburbs within the metropolitan region accounted for 32% of the jobs to which Munich area workers commuted. An astounding 56% of new commuting was to jobs outside the metropolitan region. Munich has been dispersing, despite the population gains in the central city.
Since 2001, the city of Vienna has experienced a spurt of population growth, though as in the case of London and Paris, the central city population remains well below its peak. Yet even with this increase, people continue to move out of the city of Vienna to the suburbs. The suburban and exurban ring of Vienna has gained more than 36,000 domestic migrants from Vienna and the rest of the nation since 2001. The central city of Vienna itself lost 11,000 domestic migrants. The city of Vienna has grown because of huge international migration, which alone is greater than the overall population growth of the city.
As in Vienna, the recent increase in European central city growth corresponds to increasing migration into Western Europe, especially from the eastern nations added to the European Union.
Lewyn lapses into the logical fallacy that imagines suburbanization would not have happened without high-quality roadways (freeways and motorways). Europe is the best proof that this is pure folly. As Peter Gordon and I showed, suburbanization has been similar among European urban areas, both those with and without urban freeways.
Lewyn is right that the European and American urban forms are different. Bruegmann never said otherwise. But to suggest that suburbanization in Europe is "limited" is absurd. In the United States, 75% of residents in large metropolitan areas live in the suburbs. In Europe, the number of large metropolitan area residents living in the suburbs is 65%.
Lewyn also gets it partially right on how land regulation and zoning can induce greater suburbanization than in a less regulated environment. Proof positive of this are the low density suburbs of the American Northeast, where suburban development consumed far more land because of large lot zoning. The planners required that fewer houses be built on a given amount of land, which produced far lower population densities and far greater suburbanization. If Los Angeles had suburbanized at the same density as Boston, it would cover three times as much area.
Moreover, suburbanization is the rule virtually around the world. More than 90 percent of metropolitan-area growth in the largest metropolitan areas of Japan, Australia, New Zealand and Canada has been in the suburbs since the 1960s. Suburbanization is much more than a high-income world phenomenon, as virtually all of the world's large urban areas have developed extensive suburbs. Manila now stretches into at least four provinces beyond the National Capital District. Huge swaths of suburbanization will now be found in Sao Paulo, Beijing, Jakarta, Kolkata and virtually all major urban areas, even Lagos.
All of this brings us back to what Lewyn calls the "inevitability theory." The record is clear-suburbanization, whether in Portland, Paris, Perth or the Philippines is driven by "increasing wealth," just as Bruegmann contends-people are voting with their feet and their dollars. Some planners around the world seek to limit or stop suburbanization. Indeed, it is reasonable to suggest that their efforts are prima facie evidence of suburbia's inevitability.